Some ideas for advanced ultimate stats

For a long time, ultimate players have tracked performance using incredibly simple metrics: D's, goals, and assists.  Many teams also track playing time, even differentiating between O and D points.  Recently, some more advanced apps have allowed for more advanced statistics tracking, which includes tracking every single throw (who threw it/caught it, where it was thrown/caught).  Although these can give some great information, they still lack a lot of things I think are important.  At the moment, the difficulty and time commitment required to collect these advanced stats is definitely the limiting factor (see recent 538 article), but I'm going to go through some stats that I think would be pretty cool to see in the future.  Some of them are fairly simple and could be implemented in USAU events right now, while others are a bit more in-depth and may require more volunteers (or paid employees... looking at you, AUDL/MLU).

Also, for a lot of these I basically made up names that I think sound cool and capture the basic idea of the measurement.  If you think there are better names for some of them, feel free to let me know.

So, in no particular order:

Handblocks/Footblocks

It's pretty crazy to me that nobody tracks these.  There is a pretty significant difference between a D downfield and a D on the mark, and it's also very easy to track.

Defensive Mark Efficiency

This is one that would definitely be more difficult, and would require that somebody with a trained ultimate eye be watching the game.  Basically, this would be a measure of the ability of a defender to limit break throws from the thrower.  I'm envisioning it being calculated by the equation below, which divides the number of break throws the person you are marking completed by the total throws he/she attempted.

^{(Break throws completed )}⁄_{(Total throws attempted)}

Although this wouldn't always be a perfect measure of how "good" your mark is, if aggregated over multiple games it should be a pretty valuable metric.  Note that in this case, having a smaller number would be better.  You could also just subtract this number from 1 to make it a slightly more intuitive stat, where having a higher number (closer to 1) would correlate to having a better mark.

Offensive Break Efficiency

We all know how important it can be for an offense to have potent break throwers that can change the angle and attack the whole field with their throws.  However, there still is no way to quantify the impact this can have in a standardized way.

Similar to above, offensive break efficiency would look like this:


^{(Break throws completed)}⁄_{(Total throws attempted)}

Of course, in this case, it is the number of break throws you completed divided by the total throws you attempted.

Offensive and Defensive Sky Percentage

Pretty self-explanatory; the main issue would be the subjective nature of what you count as a sky. Personally, I'd say any play where both players are close enough to each other to make contact would count as a sky.  If the offensive player drops or misreads the disc (with the defender in position to count it as a "sky" attempt), that would count as a sky for the D.

Effective Assists

I'm still not totally sure what the best way to measure something like this would be, but my thinking was that it would include goals score more than one pass away in the statistic.  So, if you throw a pass to someone who gets the assist, you might get 0.5 added to your effective assists.  Two passes away, you get 0.25 added.  Not sure if this would be useful, but it may help show what players are facilitating offense even if they aren't throwing all of the assists.

Offensive Usage (touches based or yardage based)

I think a similar statistic has been calculated before by Ultiworld, but I'll go through it anyways.  The way I see it, there are two ways to calculate Offensive Usage:

1) Based on touches
For this calculation, your offensive usage would simply be the ratio of touches you got compared to the total.  So, if an offensive point was a one-throw score, both the thrower and the receiver would have an offensive usage of 0.5 for that point.  An "expected" usage, when calculated like this, would be 1/7 for each player (although there are many other factors that would change this, like weather conditions, team strategy, player position, etc.).

2) Based on yardage
This is one that I believe has been measured before and can be measured using stat-tracking apps currently available (though I'm not positive on that - if anyone knows for sure, let me know in the comments).  For this measurement, your offensive usage stems from the percentage of the team's total yards you are responsible for (either based on throwing or receiving.   Let's imagine that a player catches the pull, then throws a 50 yard huck down the field.  The cutter then throws a 5 yard pass for a score.  The handler would have an offensive usage of 50/110, while the first cutter would have 55/110, and the last cutter (who scores the goal) would have 5/110.  Notice that you have to double the total yards covered, since you count both throwing and receiving yards.  The next level would be to separate throwing and receiving efficiencies to see which players are bigger threats downfield vs with the disc.

Another application of a metric like this would be allowing teams, coaches, and analysts to have a quantifiable measure of players' involvement with the disc.  If a team's game plan is to limit the amount of damage Beau Kittredge does downfield, a coach would be able to look at this statistic and immediately know if their team is being effective.

Defensive Usage (touches based or yardage based)

Defensive Usage would simply be the offensive usage of the opposing player while you are guarding them.  By itself it may not be super valuable, but when used with Offensive Usage, it can give you more information, such as...

Defensive Efficiency rating

This would measure the impact that a defender has on the player they are defending.  It would be the percent difference between the average offensive usage of the offensive player and the offensive usage of that player when being guarded by a particular defender (so, percent difference between offensive usage for O player and defensive usage for D player).  Because of the way this is calculated, this would be a measure of "shut-down defense", which is something that is not usually captured by keeping track of D's.

As an example: if Opi Payne's normal Offensive Usage is 20%, but her offensive usage when being guarded by Kelly Johnson is only 18%, that would give Kelly Johnson a Defensive Efficiency Rating of 10%.  You could then average Kelly's Defensive Efficiency rating for each player she guards, and get an overall measure of her shut-down defense.



There are a couple other half-ideas that I have floating around in my head, but haven't fully fleshed out yet; maybe these will be included in a future post.  For now, I'd love to hear what other people think about these stats, and any way to track them.  Personally, I'd like to try tracking some of these with my winter league team (all of our WL games are filmed and uploaded online, which would aid this process immensely).  If you'd like to try tracking some of these, I'd love to hear about it!